Compass is closing in on an FDA decision, targeting NDA submission by the end of the year, which means the scope of its patent protection now affects real near-term decisions for every company developing psilocybin. The question that becomes clear is who can compete, who can license, and who gets sued?
Compass has never patented psilocybin itself (that wouldn't be possible since it's a naturally occurring compound), only this one specific synthetic crystal form: COMP360. Compass's IP rests almost entirely on Polymorph A, a specific crystalline form of synthetic psilocybin patented in 2019, plus a related hydrate form and downstream claims on dosage and indications (TRD, MDD, and as of 2024, anxiety). The portfolio runs to roughly ten granted patents across the US, UK, Germany, and Hong Kong. Compass has kept filing and winning new ones since, most recently applications published in 2025. Every claim depends on the same narrow definition: a psilocybin crystal structure that matches specific X-ray diffraction measurements within a tolerance of just plus or minus 0.10 degrees. This is a tight enough window that psilocybin with slightly different measurements wouldn't fall under the patent.
That narrowness is the whole story. Freedom to Operate, a nonprofit set up specifically to contest psychedelic patents, challenged Compass's core patents in 2021-2022 before the Patent Trial and Appeal Board (PTAB), arguing the crystal form wasn't actually new, and that older, publicly known psilocybin samples already had it. The PTAB disagreed and let the patents stand, a decision reaffirmed on appeal in 2023. But in reaching that decision, the PTAB also said the patents only cover psilocybin matching those exact x-ray measurements, nothing broader. So if the outcome was a real win for Compass, it was a narrow one: a different crystal form, salt, or hydrate has a real argument for falling outside the claims entirely.
A separate nonprofit, Porta Sophia, has had more success working earlier in the process, submitting evidence against Compass's patent applications before they're granted rather than after. In one case, this pushed Compass to drop 137 of 162 claims (the individual assertions in a patent application, not necessarily distinct chemicals) in one pending application, including the attempt to claim the therapy room's furniture and lighting as part of the invention.
However, since late 2025, the US Patent Office has made it structurally harder to even get a hearing on a patent challenge, a policy change that cut the rate of challenges being heard by roughly 65 percent year over year. That's a new administrative hurdle before the argument even gets considered, and it favors Compass by default going forward, independent of whether a future challenger has a good case.
“Whether COMP360 gets approved should become clear early: Compass is aiming for a decision as soon as Q1 2027.”
Separately, Colorado and several other states have introduced or passed laws that legalize psilocybin at the state level only if it's “crystalline polymorph psilocybin,” language that in practice names only Compass's formulation. This is well documented (Colorado's bill passed, and Porta Sophia has publicly described lawmakers in at least one other state not realizing the bill was product-specific), though it's still an evolving situation.
If Compass's patents only cover psilocybin with that exact x-ray signature, a company using a different salt, a different crystal form, or a different molecule entirely has a real argument they simply aren't covered. That's the logic several competitors appear to be building on.
Cybin, now rebranded as Helus Pharma, is developing a deuterated version of psilocybin, meaning specific hydrogen atoms are chemically swapped for a heavier isotope, deuterium. That produces a different molecule with different physical properties (e.g. more stable) than the crystal Compass patented.
CaaMTech has built its portfolio specifically around novel psilocin salts and prodrugs, different chemical relatives of psilocybin rather than psilocybin itself, which reads as a deliberate design-around of Compass's claims.
Filament Health manufactures botanically extracted psilocybin drawn directly from mushrooms rather than synthesizing it. A company could in principle patent a novel extraction or purification process, but the extracted compound itself sits outside Compass's synthetic crystal claims entirely.
Terran Biosciences has arguably the most deliberate non-infringement position of the group. In 2023 it published its own patent application on what it describes as the first new salt and co-crystal forms of psilocybin, including psilocybin HCl forms A, B, and C, a psilocybin HCl co-crystal, psilocybin edisylate forms A and B, and psilocybin mesylate form A. Compass's Polymorph A is a specific crystal form of the free zwitterion, a psilocybin with no salt attached. Hence, Terran's forms are chemically different by design: each pairs the psilocybin molecule with a different acid to form a distinct salt. That's a meaningfully different enough chemical structure than what Compass's claims cover, not just a different manufacturing process for the same thing. Therefore, Terran's own compounds have a real basis for sitting outside Compass's crystal claims. The complication is that Compass has since filed its own pending applications reaching into salt forms, including psilocybin HCl specifically with defined dose ranges, which would put the two companies' claims in direct overlap if Compass's broader application is ever granted. Whether Terran's 2023 publication counts as evidence that blocks Compass from later patenting HCl salts, or whether Compass's claims end up narrow enough to coexist with Terran's, is not something that has been resolved publicly yet.
The one broader complication for everyone: Compass has separately tried, in pending applications, to claim much broader language like "psilocybin or an active metabolite thereof," which would reach beyond Polymorph A and could theoretically cover other companies' salts and prodrugs, including CaaMTech's and Terran's, if granted. Whether that broader language survives examination is unresolved and worth watching closely.
Whether COMP360 gets approved should become clear early: Compass is aiming for a decision as soon as Q1 2027, if its submission timeline holds.